Fuel cell hybrid electric vehicle powertrain modelling and testing

Abstract

In order to meet the 2050 targets of an 80% reduction in greenhouse gas emissions, electrification of transport is required. Of the zero-emission technologies relating to automotive applications hydrogen fuel cells, lithium-ion batteries and supercapacitors have received the greatest attention.

This thesis presents work on the development and implementation of lithium-ion battery, proton exchange membrane fuel cell and supercapacitor models with the aim of developing the tools and techniques required in assessing their feasibility in automotive applications. Experimental validation of each of these devices is provided with insight given into the physical performance limitations of each device. Analysis is then presented on overall powertrain configurations with a focus on the performance of passive hybridisation as a means of reducing the cost of a vehicle powertrain whilst retaining the advantages of hybridisation.

Four main chapters of content relating to work on: lithium-ion batteries, proton exchange membrane fuel cell, supercapacitors and vehicle system level analysis is presented with distinct conclusions and novel work presented in each chapter. Lithium-ion batteries The mathematical framework on the development of a psuedo 2D thermally-coupled electrochemical battery model is presented. This was parameterised using a genetic algorithm based technique against pulsed discharge test data for a 4.8 Ah lithium-polymer cell. This physics-based model was used to develop a means of tracking stoichiometric drift of battery electrodes using a simulated slow rate cyclic voltammetry technique as well as the development of a novel differential thermal voltammetry technique for the extraction of the same information as slow rate cyclic voltammetry but at a much faster rate. The differential voltammetry technique was then used to infer stoichiometric drift in a battery. The lithium-ion battery model was also used to investigate the scale up effects from single cell to large automotive scale packs. It was found that interconnect resistances in highly parallel packs can cause significant load inhomogeneities due to the increased overpotential caused by the interconnects which can be on the same order as the battery impedance. Cells near to the pack load points were found to experience the highest loads, with highly transient load conditions amplifying the effect. Proton exchange membrane fuel cells The mathematical framework for the development of a proton exchange membrane fuel cell model which accounted for transient thermal, mass balance and water management effects and the associated balance-of-plant system was presented. This was validated against experimental data from an in-house developed 9.5 kW 75 cell fuel cell system. Inhomogeneities in the reactant delivery, and thus performance of cells, in large automotive stacks were investigated with a focus on localised flooding leading to failure through pin-hole formation. It was shown that low pressure systems suffer from the increased risk of ooding, with location of the cell relative to the inlet/ outlets of the reactants being a critical parameter. Flooding was then shown to lead to catastrophic failure of the fuel cell stack through pin-hole formation which lead to a cascading potential instability and decay due to the bipolar coupling of the cells and anode side hydrogen cross over, respectively.